FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

  • Moritz Weißbrich
  • Guillermo Payá-Vayá
  • Lukas Gerlach
  • Holger Blume
  • A. Najafi
  • A. García-Ortiz

Externe Organisationen

  • Universität Bremen
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des Sammelwerks2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS)
UntertitelProceedings
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers Inc.
Seiten1-8
Seitenumfang8
ISBN (elektronisch)978-1-5090-6462-5
ISBN (Print)978-1-5090-6463-2
PublikationsstatusVeröffentlicht - Sept. 2017
Veranstaltung27th International Symposium on Power and Timing Modeling, Optimization and Simulation, PATMOS 2017 - Thessaloniki, Griechenland
Dauer: 25 Sept. 201727 Sept. 2017

Abstract

ASICs for Stochastic Computing conditions are designed for higher energy-efficiency or performance by sacrificing computational accuracy due to intentional circuit timing violations. To optimize the stochastic gate-level circuit behavior of a specific design, iterative timing analysis campaigns have to be carried out for a variety of chip temperature- and supply voltage-dependent timing corner cases. However, the application of common event-driven logic simulators usually leads to excessive analysis runtimes, increasing design time for hardware developers. In this paper, a gate-level netlist-oriented FPGA-based timing analysis framework is proposed, offering a runtime-configuration mechanism for emulating different timing corner cases in hardware without requiring multiple FPGA bitstreams. For an exemplary timing analysis campaign of an existing chip design, speed-up factors of up to 267 are achieved while maintaining timing behavior deviations lower than 1.05% to timing simulations.

ASJC Scopus Sachgebiete

Ziele für nachhaltige Entwicklung

Zitieren

FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework. / Weißbrich, Moritz; Payá-Vayá, Guillermo; Gerlach, Lukas et al.
2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS): Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017. S. 1-8.

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Weißbrich, M, Payá-Vayá, G, Gerlach, L, Blume, H, Najafi, A & García-Ortiz, A 2017, FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework. in 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS): Proceedings. Institute of Electrical and Electronics Engineers Inc., S. 1-8, 27th International Symposium on Power and Timing Modeling, Optimization and Simulation, PATMOS 2017, Thessaloniki, Griechenland, 25 Sept. 2017. https://doi.org/10.1109/PATMOS.2017.8106956
Weißbrich, M., Payá-Vayá, G., Gerlach, L., Blume, H., Najafi, A., & García-Ortiz, A. (2017). FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework. In 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS): Proceedings (S. 1-8). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PATMOS.2017.8106956
Weißbrich M, Payá-Vayá G, Gerlach L, Blume H, Najafi A, García-Ortiz A. FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework. in 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS): Proceedings. Institute of Electrical and Electronics Engineers Inc. 2017. S. 1-8 doi: 10.1109/PATMOS.2017.8106956
Weißbrich, Moritz ; Payá-Vayá, Guillermo ; Gerlach, Lukas et al. / FLINT+ : A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework. 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS): Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017. S. 1-8
Download
@inproceedings{4ce9e1b993e6434798dba00681b319af,
title = "FLINT+: A Runtime-Configurable Emulation-Based Stochastic Timing Analysis Framework",
abstract = "ASICs for Stochastic Computing conditions are designed for higher energy-efficiency or performance by sacrificing computational accuracy due to intentional circuit timing violations. To optimize the stochastic gate-level circuit behavior of a specific design, iterative timing analysis campaigns have to be carried out for a variety of chip temperature- and supply voltage-dependent timing corner cases. However, the application of common event-driven logic simulators usually leads to excessive analysis runtimes, increasing design time for hardware developers. In this paper, a gate-level netlist-oriented FPGA-based timing analysis framework is proposed, offering a runtime-configuration mechanism for emulating different timing corner cases in hardware without requiring multiple FPGA bitstreams. For an exemplary timing analysis campaign of an existing chip design, speed-up factors of up to 267 are achieved while maintaining timing behavior deviations lower than 1.05% to timing simulations.",
author = "Moritz Wei{\ss}brich and Guillermo Pay{\'a}-Vay{\'a} and Lukas Gerlach and Holger Blume and A. Najafi and A. Garc{\'i}a-Ortiz",
note = "Funding information: This work was partly funded by the German Research Council (DFG) under project number PA 2762/1-1.; 27th International Symposium on Power and Timing Modeling, Optimization and Simulation, PATMOS 2017 ; Conference date: 25-09-2017 Through 27-09-2017",
year = "2017",
month = sep,
doi = "10.1109/PATMOS.2017.8106956",
language = "English",
isbn = "978-1-5090-6463-2",
pages = "1--8",
booktitle = "2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS)",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

Download

TY - GEN

T1 - FLINT+

T2 - 27th International Symposium on Power and Timing Modeling, Optimization and Simulation, PATMOS 2017

AU - Weißbrich, Moritz

AU - Payá-Vayá, Guillermo

AU - Gerlach, Lukas

AU - Blume, Holger

AU - Najafi, A.

AU - García-Ortiz, A.

N1 - Funding information: This work was partly funded by the German Research Council (DFG) under project number PA 2762/1-1.

PY - 2017/9

Y1 - 2017/9

N2 - ASICs for Stochastic Computing conditions are designed for higher energy-efficiency or performance by sacrificing computational accuracy due to intentional circuit timing violations. To optimize the stochastic gate-level circuit behavior of a specific design, iterative timing analysis campaigns have to be carried out for a variety of chip temperature- and supply voltage-dependent timing corner cases. However, the application of common event-driven logic simulators usually leads to excessive analysis runtimes, increasing design time for hardware developers. In this paper, a gate-level netlist-oriented FPGA-based timing analysis framework is proposed, offering a runtime-configuration mechanism for emulating different timing corner cases in hardware without requiring multiple FPGA bitstreams. For an exemplary timing analysis campaign of an existing chip design, speed-up factors of up to 267 are achieved while maintaining timing behavior deviations lower than 1.05% to timing simulations.

AB - ASICs for Stochastic Computing conditions are designed for higher energy-efficiency or performance by sacrificing computational accuracy due to intentional circuit timing violations. To optimize the stochastic gate-level circuit behavior of a specific design, iterative timing analysis campaigns have to be carried out for a variety of chip temperature- and supply voltage-dependent timing corner cases. However, the application of common event-driven logic simulators usually leads to excessive analysis runtimes, increasing design time for hardware developers. In this paper, a gate-level netlist-oriented FPGA-based timing analysis framework is proposed, offering a runtime-configuration mechanism for emulating different timing corner cases in hardware without requiring multiple FPGA bitstreams. For an exemplary timing analysis campaign of an existing chip design, speed-up factors of up to 267 are achieved while maintaining timing behavior deviations lower than 1.05% to timing simulations.

UR - http://www.scopus.com/inward/record.url?scp=85043467254&partnerID=8YFLogxK

U2 - 10.1109/PATMOS.2017.8106956

DO - 10.1109/PATMOS.2017.8106956

M3 - Conference contribution

AN - SCOPUS:85043467254

SN - 978-1-5090-6463-2

SP - 1

EP - 8

BT - 2017 27th International Symposium on Power and Timing Modeling, Optimization and Simulation (PATMOS)

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 25 September 2017 through 27 September 2017

ER -

Von denselben Autoren